Stepanov E.P.

The traffic load balancing problem is relevant in modern networks that have many alternative routes between any pair of nodes. Balancing provides a uniform load of network resources. The paper proposes a method for adaptive queueing policy configuration on a switch to get a uniform queue load on the switch output ports. Because modern applications require the data transmission delay to be around milliseconds, reinforcement learning method DQN was applied to solve the problem. An experimental study demonstrated the convergence of the proposed method during the training to uniform queue load at output ports.

The problem of applying erasure coding methods at the transport level to restore lost packets is considered. This will allow to avoid multiple retransmissions of the same packet, reduce data transmission delays, and waste of network resources. The basic idea behind erasure coding methods is to introduce redundancy into the transmitted data, which will allow the lost data to be recovered on the receiver side. The paper considers various erasure coding methods at the transport level, selects the most promising ones based on the computational complexity of the encoding and decoding algorithms, as well as the effect of redundancy on the data transmission delay and the transport connection loss level. The required level of redundancy in the selected error-correcting coding methods is given, depending on the requirements for the loss level in the transport connection and its quality characteristics.